beta-Catenin gene mutations and microsatellite instability (MI) have been reported in endometrioid ovarian carcinomas. In colon but not endometrial cancer, beta-catenin gene mutations are associated with a replication error phenotype and MI. In this study the authors investigate whether beta-catenin mutations and MI are two independent oncogenic pathways in endometrioid ovarian carcinomas. They also evaluate the usefulness of these molecular markers in determining the primary origin of simultaneous tumors in the ovary and endometrium. This study was performed on 26 patients diagnosed with primary endometrioid ovarian carcinoma, five of whom also had pathologically diagnosed primary synchronous endometrioid endometrial carcinoma. Immunohistochemical and molecular analyses indicated that there were 25 primary ovarian tumors with four primary synchronous endometrial cancers and one ovarian metastasis of a primary endometrial carcinoma. All studies were performed on formalin-fixed, paraffin-embedded tissue samples. The beta-catenin expression pattern (nuclear vs. membranous) was analyzed immunohistochemically. Mutations in exon 3 of the beta-catenin gene were studied by polymerase chain reaction, single-strand conformational polymorphism, and direct sequencing. MI status was established by studying BAT-26 and BAT-25 mononucleotide repeats. In the group with 21 single ovarian tumors, 18 (85%) had beta-catenin nuclear expression, eight (38%) had beta-catenin gene mutations (always associated with beta-catenin nuclear expression), and four (19%) had MI. Only one case (5%) had both beta-catenin gene mutations and MI. The mutations affected one of the serine/threonine residues targeted for phosphorylation by glycogen synthase kinase-3beta or adjacent residues. At codon 32, a GAC-to-TAC (D32Y) change was found; at codon 33, two TCT-to-TGT (S33C) changes were found; at codon 37, three TCT-to-TTT (S37F) changes and one TCT-to-TGT (S37C) change were found; and, lastly, one ACC-to-GCC change at codon 41 (T41A) was detected. Four of the 25 endometrioid ovarian carcinomas (16%) had an associated synchronous endometrial carcinoma. There was a higher percentage of beta-catenin mutations (n = 3, 75%) in synchronous ovarian carcinomas than in single ones, although with a similar percentage of MI (n = 1, 25%). beta-catenin mutations were S37C in two cases and D32G in one. One of the four endometrial carcinomas showed an S33C beta-catenin mutation, and two carcinomas had MI. None of the four tumors had both beta-catenin gene mutation and MI. beta-catenin gene mutations were always associated with a nuclear beta-catenin expression pattern, whereas MI was associated with a membranous pattern. In one patient both the ovarian and the endometrial carcinomas had beta-catenin gene mutations, in another patient both tumors showed MI, whereas in the remaining two patients the ovarian carcinomas showed beta-catenin gene mutations and the endometrial carcinomas showed MI. To summarize, the results of this study suggest that beta-catenin mutations and MI could represent two independent pathways in endometrioid ovarian carcinomas because they occur simultaneously very infrequently (in 5% of these cases). beta-catenin mutations are always associated with a nuclear beta-catenin expression pattern, whereas cases with a replication error -plus phenotype showed no abnormal beta-catenin subcellular localization. The study of the beta-catenin expression pattern, beta-catenin mutations, and MI, together with conventional clinicopathologic findings, could aid in distinguishing between the metastatic or independent origin of simultaneous endometrioid ovarian and endometrial carcinomas. Tumors with identical immunohistochemical and molecular features should therefore be considered to have a common origin.